Tambu Insights Consulting & Women in Digital Transformation Foundation

Success cannot be measured by the quality of the conversation alone. It must be measured by what the conversation commits to. Three outcomes would make Geneva a genuine milestone. First, a Co-Chairs' Summary that names consumer redress as a foundational user right. AI systems are already making consequential decisions about people's credit, employment, and safety across the Global South, and most affected communities have no right to explanation, no appeal mechanism, and no recourse. A successful Dialogue names this gap explicitly and calls for accessible redress mechanisms globally, not only in jurisdictions with enforcement capacity. Second, a public implementation scorecard: a set of traceable commitments reviewed at the opening of the May 2027 New York session. The credibility of this Dialogue depends on whether July produces measurable follow-through, not declarations that disappear between editions. Third, meaningful civil society participation, not as observers reacting to government statements, but as structured contributors within each thematic session. The communities most affected by AI governance failures are disproportionately represented by civil society organizations. Their field evidence belongs in the room, not in the margins. A successful Dialogue also extends existing ITU and UNESCO fellowship and support mechanisms explicitly to civil society participants for 2027, ensuring that inclusion is not contingent on organizational budget size. Finally, success means the Dialogue produces language the Independent International Scientific Panel can hold governments accountable to, connecting the scientific evidence base directly to governance commitments. If Geneva leaves without traceable, specific, and inclusive outcomes, it will have produced a process, not progress.

• AI capacity-building
• Protection and promotion of human rights
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Transparency, accountability, and human oversight

These four priorities reflect the lived realities of the communities our work serves across Africa and the Global South. AI capacity-building is foundational, but must be reframed beyond infrastructure. Real capacity means women and marginalized communities understanding their rights as AI users, data subjects, and contributors, not simply gaining access to AI tools designed elsewhere. At WinDT Foundation, our AI literacy and safety programming demonstrates that rights awareness is as critical as technical skills. Protection and promotion of human rights is urgent and concrete. AI systems are already causing measurable human rights harms, including algorithmic hiring discrimination, biased credit scoring, and AI-enabled gender-based violence such as deepfakes and surveillance, with no redress mechanisms in most of the Global South. In October 2025, France's equality regulator ruled Facebook's job ad algorithm discriminatory and sexist in a jurisdiction with enforcement. In most of the Global South, there is none. This is not a future risk. It is a present reality requiring universal governance now. The social, economic, ethical, cultural, linguistic and technical implications of AI captures the breadth of impact our consulting work at Tambu Insights addresses, from national AI strategy advisory to MEL frameworks assessing digital ecosystem outcomes across nineteen countries. The economic implications for women as workers, entrepreneurs, and tool users are particularly underexamined and deserve dedicated attention in Geneva. Transparency, accountability, and human oversight are the necessary conditions for the other three to be actionable. Without mandatory human rights impact or community impact assessments before deployment, without the right to explanation, and without meaningful human oversight in high-stakes decisions, the other priorities remain aspirational. The UN High Commissioner for Human Rights called for human rights impact assessments as a matter of urgency at the AI Impact Summit in February 2026. Geneva should operationalize that call.

Two significant gaps deserve attention. First, consumer redress and user rights. Reading the four proposed thematic clusters carefully, not one explicitly names the right to redress, the right to explanation, or any user protection mechanism. This is a critical omission. Across the Global South, individuals subject to consequential AI decisions, including loan denials, job screening, benefit eligibility, and criminal risk scoring, have no meaningful recourse. Consumer redress is not a subset of human rights or transparency. It is a distinct, operational issue requiring its own governance architecture, including accessible complaint mechanisms, burden-of-proof standards, and cross-border enforcement cooperation. Second, data governance and benefit-sharing. The communities whose data trains AI systems, often low-income, often in the Global South, and disproportionately women, share in neither the governance nor the economic benefit of those systems. This creates a structural inequity that none of the seven thematic areas directly addresses. Data sovereignty, community consent, and benefit-sharing frameworks should be named as cross-cutting priorities, particularly as AI developers increasingly rely on data extracted from communities with limited awareness of how it is used. Both gaps have a gender dimension that deserves explicit recognition. Women are disproportionately subject to AI-enabled harms, including deepfakes, algorithmic surveillance, and gender-biased automated decisions in employment and financial services, and are underrepresented in AI development, governance, and standard-setting processes. Meaningful inclusion requires more than access to AI tools. It requires women's participation as designers, decision-makers, and rights-holders throughout the AI lifecycle. These issues cut across all seven listed themes and should be treated as horizontal priorities shaping how each theme is approached in Geneva.

Across sub-Saharan Africa, the governance gaps in the four priority areas are not abstract. They are producing measurable harm today, particularly for women and underserved communities. On human rights and accountability, AI-driven financial services are expanding rapidly across the region while governance frameworks lag far behind. Research published in 2025 found that African fintech credit scoring algorithms impose a 37 percent underfunding penalty on women-led small businesses, driven by hidden proxies including sector misclassification and network bias that favors male-dominated affiliations. In Kenya's digital lending sector, women with limited mobile and internet access receive lower credit scores because behavioral data used in scoring underrepresents their financial activity. These are human rights failures with direct economic consequences, and no regional redress mechanism exists. On capacity-building, the gap is not only technical. Through our AI literacy programming, we consistently find that women across Africa have little or no awareness that AI systems are making decisions about their access to credit, employment, and services. Rights awareness is not part of any current capacity-building framework at scale. This must change. On transparency and human oversight, most AI systems deployed in high-stakes sectors across the region, including mobile lending, agricultural insurance, and public service delivery, operate as black boxes. Users have no right to explanation and no pathway to challenge decisions. The EU AI Act classifies credit scoring as high-risk and mandates transparency and human oversight. No equivalent standard exists across most of Africa. African governments are actively developing national AI strategies and civil society organizations are building community-level awareness. The Global Dialogue can strengthen these efforts by establishing minimum standards for user rights, redress, and transparency that are realistic for low-resource contexts.

The AI Dialogue occupies a position no existing mechanism holds: a universal, inclusive United Nations platform where every country, including those largely absent from G7, G20, OECD, and Council of Europe processes, has an equal seat. That is its most important and distinctive role. Three specific contributions the Dialogue is positioned to make. First, it can serve as an interoperability anchor. AI governance is currently fragmented across regional frameworks, national laws, and voluntary principles. The Dialogue cannot and should not replace these efforts. But it can identify where they diverge, where they overlap, and where minimum common standards are achievable, particularly on user rights, redress mechanisms, and human rights impact assessments. Second, it can rebalance whose priorities shape global norms. Current governance frameworks reflect the priorities of technologically advanced economies. Sub-Saharan Africa, small island developing states, and least developed countries face AI-driven harms today, including biased credit scoring, algorithmic targeting, and AI-enabled gender-based violence, without governance frameworks adequate to address them. The Dialogue can ensure their experience informs global standards, not the other way around. Third, it can close the accountability gap between declarations and implementation. The Dialogue meets annually, alternating between Geneva and New York. That cadence creates a structural opportunity to track whether commitments made in one session are honored before the next. A public implementation scorecard tied to each Co-Chairs' Summary would give this Dialogue a degree of accountability that most comparable processes lack. The Dialogue will not solve fragmentation alone. But governed well, it can be the place where the Global South is not just consulted but genuinely shapes the norms that govern AI systems already operating in their communities.

Several existing mechanisms provide a strong foundation the Dialogue should explicitly build upon rather than duplicate. UNESCO's Recommendation on the Ethics of AI, adopted by 194 member states, and its Readiness Assessment Methodology, now used by over 75 countries, provide the most widely adopted normative and diagnostic framework available. The Dialogue should treat these as a baseline, not a parallel track. The OECD AI Principles and Policy Observatory provide the most comprehensive evidence base on AI governance globally and have been explicitly linked to UN mechanisms through enhanced OECD-UN collaboration announced in 2024. The Dialogue should draw on this evidence base directly, particularly for capacity-building discussions. The African Union Continental AI Strategy and the AU-EU digital partnership represent emerging regional frameworks that deserve explicit recognition. Africa is moving from bystander to active participant in global AI governance. The Dialogue should accelerate that shift, not treat it as a side conversation. The UNESCO-UNDP joint capacity-building initiative on data governance for inclusive digital futures, hosted at the ITU Academy and supported by the EU Global Gateway, brings together government officials from across Africa, Asia, and the Arab States. This practical model of South-South and triangular cooperation should be recognized and scaled within the Dialogue's capacity-building workstream. The added value the Dialogue brings is universality and accountability. Existing mechanisms are either geographically limited, voluntary, or disconnected from implementation review. The Dialogue can be the place where these efforts are connected, their gaps are named, and their commitments are tracked, with the Independent International Scientific Panel providing the evidence base to hold all parties accountable.

Meaningful multistakeholder participation requires structural design, not goodwill alone. Each stakeholder group brings a distinct contribution that the Dialogue's format must actively enable. Governments bring mandate, resources, and the authority to translate commitments into law. Their role is strongest when they come prepared with specific national positions, not general statements, and when they engage across sessions rather than delivering standalone interventions. Civil society organizations bring field evidence from the communities most affected by AI governance failures. Their contribution is only realized if the Dialogue assigns them structured time within each thematic session, not reserved only for opening or closing plenaries. The joint civil society letter of April 2026 called for a minimum of 50 percent rotating time for non-governmental stakeholders. This recommendation deserves serious consideration. The private sector brings technical expertise and real-world deployment experience. Their participation is valuable but requires guardrails. Commercial interests should be disclosed transparently, and private sector contributions should be subject to the same scrutiny as other inputs. Academia and the technical community connect the Dialogue to the Independent International Scientific Panel's evidence base. Structured opportunities for Panel members to engage directly with thematic discussions, not only through formal presentations, would strengthen this link. On format, the Dialogue should move away from sequential statement-making toward moderated exchanges in which stakeholders respond to each other in real time. The launch event in September 2025 modeled this by interspersing non-governmental statements among Member State interventions. Geneva should build on that precedent. Finally, the Dialogue should publish a structured synthesis of written inputs before each session, so in-room discussions build on what has already been submitted rather than repeating it.

The most significant underrepresentation in global AI governance falls into four categories. First, women and gender-diverse people. Despite being disproportionately affected by AI-enabled harms, including deepfakes, algorithmic discrimination in hiring and credit, and surveillance, women remain underrepresented as designers, decision-makers, and rights-holders in AI governance processes. UNESCO's Women for Ethical AI platform is a positive step, but targeted inclusion of women practitioners from the Global South, not only researchers from established institutions, is still insufficient. The Women in Digital Transformation Foundation's work delivering AI literacy and rights awareness across Africa demonstrates that community-level practitioners carry evidence and perspectives that formal governance processes do not yet systematically reach. Second, communities in least developed countries and small island developing states. These communities are already subject to AI systems in financial services, agricultural advisory tools, and public administration, yet have almost no representation in the processes that govern them. Third, grassroots civil society organizations working at community level. International NGOs with UN accreditation are not a substitute for local organizations with direct community accountability. The Dialogue should create accessible virtual participation pathways with interpretation support, and recognize written submissions from community-level organizations as substantive inputs to thematic discussions. Fourth, youth. AI systems will shape the longest portion of their lives, yet youth voices remain largely absent from governance architecture. Practical steps include virtual participation with multilingual support, pre-session written input synthesis, civil society liaisons to the Co-Chairs, and a dedicated fellowship mechanism using existing ITU and UNESCO frameworks to support in-person participation from underrepresented regions in 2027. Tambu Insights Consulting and the Women in Digital Transformation Foundation welcome opportunities to facilitate and contribute to inclusion-focused workstreams within the Dialogue process.

The most effective engagement formats are those that produce outputs, not just exchanges. Three approaches deserve serious consideration. First, structured evidence sessions tied to the Independent International Scientific Panel. Rather than presenting Panel findings as a one-way briefing, each thematic discussion should open with a fifteen-minute evidence anchor from Panel members, followed by a moderated exchange in which governments and civil society respond directly to the evidence. This would ground political discussions in scientific findings and reduce the risk of repetitive general statements. Second, regional input consolidation before each session. The Dialogue should commission structured synthesis reports from each UN regional grouping, drawing on civil society, government, and academic submissions, and make these publicly available two weeks before Geneva. This allows in-room discussion to build on what has already been contributed rather than starting from scratch. The HLPF's Voluntary National Reviews offer a partial model, though a lighter and faster format is more appropriate here. Third, problem-based working groups between sessions. Rather than limiting engagement to the two-day annual meeting, the Dialogue should establish lightweight inter-sessional working groups on specific governance problems, such as consumer redress mechanisms or data benefit-sharing frameworks, with outputs feeding directly into the following year's Co-Chairs' Summary. The Internet Governance Forum's dynamic coalitions provide a precedent for this model. On virtual engagement, hybrid participation must be treated as a first-class format, not a fallback. This means dedicated moderation for virtual participants, real-time interpretation, and structured virtual queuing systems that give online speakers equivalent access to the floor as in-person participants. These formats would make the Dialogue genuinely interactive, evidence-based, and continuous rather than episodic.

Several concrete practices and approaches offer lessons the Global Dialogue should draw upon. The UNESCO Readiness Assessment Methodology is the most widely deployed diagnostic tool for AI governance capacity, used by over 75 countries across five continents as of October 2025. Its value lies in being government-led, contextually adaptable, and directly linked to actionable policy recommendations. The Dialogue should recommend its use as a baseline assessment for all member states, particularly those without existing national AI strategies. The African Union Continental AI Strategy, currently in Phase 1 implementation through 2026, represents a regional governance model built explicitly on African values and on the recognition that AI governance frameworks must reflect local realities, not simply import external standards. The AU's approach of co-designing governance frameworks with diverse stakeholders offers a model worth replicating in the Dialogue's own processes. The Women in Digital Transformation Foundation's community-based AI literacy model demonstrates that governance impact requires investment in rights awareness at the user level, not only at the policy level. Women who understand their rights as AI users, data subjects, and contributors are better positioned to hold institutions accountable. This bottom-up approach complements top-down regulatory frameworks and should be recognized in the Dialogue's capacity-building workstream. The UNESCO-UNDP joint initiative on data governance for inclusive digital futures, hosted at the ITU Academy and supported by the EU Global Gateway, brings together government officials from Africa, Asia, and the Arab States. Scaling this model within the Dialogue's continuity framework would extend its reach significantly. Tambu Insights Consulting is available to contribute advisory expertise to any of these workstreams. Dr. Anong's work spans digital economy strategy, AI governance frameworks, cybersecurity policy, gender and inclusion, and monitoring, evaluation and learning across Africa and the Global South.